NATURE 



[May 7, 1891 



subject ; and the first volume commences with an inquiry 

 into the relationship of the Neuropteroid insects of the 

 North American Carboniferous to the existing Neuroptera, 

 as exemplified by the structure of their wings. It would 

 be out of place here to allude to the variations in the 

 structure of the veins of the wings presented by different 

 groups of insects, and their derivation from a common 

 plan of structure ; and we may accordingly proceed to 

 notice the most interesting chapter in the whole volume. 

 This is the essay on Palasodictyoptera, commencing on 

 p. 283. Here we have a detailed account of the reasons 

 which induced the author to separate the whole of the 

 Palasozoic insects from the existing orders under the 

 name of Palasodictyoptera — a term first proposed by 

 Goldenberg in lieu of Dohrn's preoccupied Dictyoptera, 

 which had been suggested for an order typified by the 

 Permian Etigereon. This order is defined more by the 

 generalized characters of its various members, and the 

 lack of those special characteristics which are the pro- 

 perty of existing orders, than by any definite peculiarities 

 of its own. One of its most important features is, how- 

 ever, that the two pairs of wings are always closely 

 similar to one another, being equally membranous, and 

 with the six principal veins always developed. With the 

 exception of a i&vi cockroach-like insects found in the 

 American Trias, the Palasodictyoptera not only includes 

 all the insects of the Palaeozoic, but is restricted to that 

 period, and is, therefore, extremely convenient to the geo- 

 logist. The order is divided into various sections, which are 

 severally regarded as the ancestors of the existing orders 

 whose names they bear. Thus, the Palaeozoic cock- 

 roaches constitute the Orthopteroid Palasodictyoptera ; 

 while we have a Neuropteroid section represented by 

 Platephemera, Mia?nia, &c. ; and an Hemipteroid one 

 by the above-mentioned Eugereon, The presence in 

 wood of Carboniferous age of borings similar to those 

 made by modern Coleoptera, further suggests the exist- 

 ence of a Coleopteroid section of the order. The author 

 (p. 320) considers that such Coleopteroids "at first showed 

 no greater distinction between the front and hind wings 

 than existed in other Palaeodictyoptera ; but afterwards 

 those races were preserved in which the thickening of the 

 membrane of the upper wings the better protected the 

 insects in their burrows for the marriage flight in open 

 air." 



The author gives a still fuller account of the reasons 

 for adopting the order Palaeodictyoptera, in the essay on 

 " Winged Insects from a Palaeontological Point of View " 

 (P- 317)5 from which the preceding extract is taken. 

 Great stress is there laid on the fact that the differentia- 

 tion of wing-structure characteristic of modern insects did 

 not exist in those of Palaeozoic times ; all of them having 

 a common type of neuration barely admitting of division 

 into families. The differences in the organs of the 

 mouth, as exemplified by the biting Progonoblattina (a 

 PaL^ozoic cockroach) and the suctorial Eugereon, are 

 considered merely as physiological adaptations of no 

 morphological value (pp. 284, 285). 



The facts and arguments detailed by the author leave, 

 then, no doubt as to the close affinities and undiffer- 

 entiated characters of all the Palaeozoic insects ; and also 

 that the group Palaeodictyoptera includes the ancestors 

 of a considerable number of the existing orders of insects. 

 NO. 1 123, VOL. 44] 



Since, however, all the latter are clearly divergent 

 branches from one or more common stocks, and are in 

 no sense ancestral to one another, the suggestion arises 

 whether it might not be advisable to group all the existing 

 orders together— say, under the name of the Neodictyo- 

 pterine " series " ; and to rank the Palceodictyoptera as a 

 " series " of equal value, in which the various members 

 were not sufficiently differentiated from one another ta 

 constitute "orders." It is a very significant fact that,^ 

 while the Palaeozoic insects show ancestral forms of those 

 recent orders grouped together by Packard as the 

 Heterometabola, they include no ancestral types of the 

 more specialized orders — Lepidoptera, Hymenoptera, and 

 Diptera — constituting the Metabola. We have, therefore, 

 proof that these specialized types are of later date ; 

 and it thus appears that palaeontological evidence is in 

 favour of Packard's classification.^ Of the existing orders 

 of insects it appears, indeed, that while the Neuroptera,. 

 Orthoptera, and Coleoptera are more or less fully repre- 

 sented in the Trias, it is not till the Lias that we meet 

 with Hemiptera (Rhynchota), although Eugereoft may be 

 taken as sufficient evidence that a Triassic member of 

 that order must have existed. None of the Meta- 

 bola are known before the Lias, the Diptera and 

 Hymenoptera dating from that epoch, while the Lepido- 

 ptera are unknown till the Middle Jurassic. 



Though space does not permit of much further reference 

 to the true insects of the pre-Tertiary epochs, we cannot 

 pass over the interesting essay (p. 323) on the oldest 

 known insect larvae. These larvae, which appear to be 

 very abundant in the Trias of the Connecticut River,, 

 are known as Mormolucoides {Palephemera), and there 

 has been much discussion as to whether they indicate 

 Coleopteroid or Neuropteroid insects. Mr. Scudder's 

 mode of treating this difficult question is a model of 

 palteontological induction. After carefully reviewing 

 all the evidence, he concludes that the fossils come 

 nearer to the larvae of the Neuropterous families Per- 

 lidce, EphemeridcE, and Sialidce, and that the relationship 

 is nearest to the latter family, which belongs to the true 

 Neuroptera. Anotherexceedingiyinteresting article (p. 433) 

 refers to the cockroaches of the Fairplay beds, Colorado. 

 Several of the species from these beds belong to the 

 Palaeodictyoptera, showing the complete interdependence 

 of two of the veins of the fore-wing characteristic of the 

 Palaeozoic types. Others, however, are true Orthopteroid 

 •cockroaches, and we thus seem to have presented to our 

 view the very period when the Palaeodictyoptera were 

 passing into the Orthoptera. From the mingled Palaeo- 

 zoic and Mesozoic facies presented by their insect fauna, 

 the author is disposed to refer the Fairplay beds to the 

 Trias ; although, as is so frequently the case, the plant- 

 evidence does not accord with that presented by the 

 animals. 



Passing to the Palaeozoic Myriopods, we notice that 

 while all the forms described in the earlier essays are 

 clearly referable to extinct ordinal groups, the progress 

 of discovery has recently shown (p. 393) that side by side 

 with these lost types there existed in the Coal-measures 

 of Illinois Centipedes closely allied to existing forms, and 



^ Many authorities, attaching more importance to the nature of the meta- 

 morphosis, transfer the Coleoptera to the higher group (Holometabola). la 

 which some also include the true Neuroptera, placing the Pseudjneuroptera 

 with the Orthoptera. 



